Business Impact (ROI)

Sooner or later someone asks: "what is this tool actually getting us?" The Business Impact page is the answer — and it's built to survive a skeptical audience, because it's grounded in your own repository's history rather than a vendor's marketing claims.

The page splits honestly into two kinds of numbers: things that actually happened in your repo, and a projection whose assumptions you control and can see.

Proven impact — what actually happened

• Score validation lift. The blunt question about any risk score is does it predict anything? This stat answers it with your own bugs: how much more often did High/Critical PRs later turn out to be the origin of a bugfix, compared to Low/Moderate ones? A lift of 3× means risky-scored PRs became bug sources three times as often — the score is pointing somewhere real. A lift near 1× would mean it isn't (and we'd rather show you that than hide it — a metric you can't falsify is a metric you can't trust).
• Bugfixes traced to risky PRs — the count of real incidents whose origin scored High or Critical. Each one is a receipt: a concrete bug the score flagged in advance.
• High-risk PRs merged without tests — risk that shipped unmitigated. This is the "cost of ignoring the advice" column, useful when arguing for more testing time.

Projected savings — an estimate you control

• Estimated cost avoided multiplies three things: high-risk PRs caught × person-hours of debugging each one typically represents × your hourly rate. The person-hours and hourly-rate inputs are editable, on purpose: plug in your team's real numbers and the estimate is yours, not ours. It's an estimate either way — but one whose assumptions are visible instead of buried.
• Future Debug Cost now vs. baseline — how far average risk has moved since you started.
• Analysis coverage — the share of your PRs that have actually been analyzed. If this is low, every other number on the page understates reality; a sync fixes it.

Trends

The charts give the story shape over time: average Future Debug Cost against your baseline, cumulative estimated cost avoided, and how the risk mix (critical / high / moderate / low) has shifted.

Using it

This is the page to screen-share with a stakeholder. The honest pitch it supports: "here's evidence the score predicts our real bugs, here's what shipped risky anyway, and here's a savings estimate using our own numbers." Skeptics tend to respect the parts you didn't inflate.